Compare the Decrease in Visceral Adipose Tissue in People with Obesity and Prediabetes vs. Obesity and Type 2 Diabetes Treated with Liraglutide

Abstract

Obesity is considered a global pandemic. In Mexico, 7/10 adults, 4/10 adolescents, and 1/3 children are overweight or obese, and it is estimated that 90% of cases of type 2 diabetes (T2D) are attributable to these pathologies. Visceral adipose tissue (VAT) presents increased lipolysis, lower insulin sensitivity, and greater metabolic alterations. Glucagon-like peptide-1 (GLP-1) is a polypeptide incretin hormone that stimulates insulin secretion dependent on the amount of oral glucose consumed, reduces plasma glucagon concentrations, slows gastric emptying, suppresses appetite, improves insulin synthesis and secretion, and increases the sensitivity of β cells to glucose. Liraglutide is a synthetic GLP-1 analog that reduces VAT and improves the expression of Glucose transporter receptor type 4 (GLUT 4R), Mitogen-activated protein (MAP kinases), decreases Fibroblast growth factor type β (TGF-β), reactivates the peroxisome proliferator-activated receptor type ɣ (PPAR-ɣ) pathway, and decreases chronic inflammation. Currently, there are many studies that explain the decrease in VAT with these medications, but there are no studies that compare the decrease in patients with obesity and prediabetes vs. obesity and type 2 diabetes to know which population obtains a greater benefit from treatment with this pharmacological group; this is the reason for this study. The primary objective was to compare the difference in the determination of visceral adipose tissue in people with obesity and type 2 diabetes vs. obesity and prediabetes treated with liraglutide. Methods: A quasi-experimental, analytical, prolective, non-randomized, non-blinded study was conducted over a period of 6 months in a tertiary care center. A total of 36 participants were divided into two arms; group 1 (G1: Obesity and prediabetes) and group 2 (G2: Obesity and type 2 diabetes) for 6 months. Inclusion criteria: men and women ≥18 years with type 2 diabetes, prediabetes, and obesity. Exclusion criteria: Glomerular filtration rate (GFR) < 60 mL/min/1.73 m² elevated transaminases (>5 times the upper limit of normal), and use of non-weight-modifying antidiabetic agents. Conclusions: No statistically significant difference was found in the decrease in visceral adipose tissue when comparing G1 (OB and PD) with G2 (OB and T2D). When comparing intragroup in G2 (OB and T2D), greater weight loss was found [(−3.78 kg; p = 0.012) vs. (−3.78 kg; p = 0.012)], as well differences in waist circumference [(−3.9 cm; p = 0.049) vs. (−3.09 cm; p = 0.017)], and glucose levels [(−1.75 mmol/L; p = 0.002) vs. (−0.56 mmol/L; p = 0.002)], A1c% [(−1.15%; p = 0.001) vs. (−0.5%; p = 0.000)].

1. Introduction

Adipose tissue (AT) is an endocrine organ characterized by the secretion of multiple hormones that modify insulin and glucose function and cardiovascular risk. There are three main entities that are interrelated that cause dysglycemia: obesity, prediabetes, and type 2 diabetes.

Obesity (OB) is considered a global pandemic. The World Health Organization (WHO) defines it as excess adipose tissue, which predisposes to the development of type 2 diabetes (T2D) [1]. According to its figures, in 2019 there were 650 million adults over 18 years of age with OB [2]. In Mexico, the prevalence increased from 35.6% to 42.2% between 2012 and 2018 [3]. The importance of excess adipose tissue lies in the increased risk of developing T2D and the increase in body mass index (BMI) [4].

Type 2 diabetes (T2D) is a chronic disease that occurs when the pancreas does not secrete enough insulin or when the body does not use it efficiently [5]. According to the International Diabetes Federation (IDF), there are 537 million adults aged 20 to 79 years living with diabetes worldwide, and in North America, 1 in 7 adults (51 million) and 1 in 4 adults live with undiagnosed diabetes [6]. In Mexico, according to the National Health and Nutrition Survey (ENSANUT), 8.6 million people have been diagnosed with T2D, and it is estimated that by 2045, 20% of the Mexican population could develop T2D [7]. According to the American Diabetes Association (ADA) 2023, the diagnostic criteria for T2DM are fasting glucose ≥ 126 mg/dL, glycated hemoglobin (A1c%) ≥ 6.5%, 2 h postprandial glucose ≥ 200 mg/dL, and Symptoms + Random glucose ≥ 200 mg/dL [8].

Prediabetes (PD) is a condition with glucose levels above normal without meeting the criteria for diabetes [9]. According to the ADA 2023, the criteria for PD are fasting glucose ≥ 100–125 mg/dL, A1c ≥ 5.7–6.4%, and 2 h postprandial glucose ≥ 140–199 mg/dL [8]. The importance of PD lies in its importance as a precursor to the development of type 2 diabetes. Worldwide, there are 88 million people over the age of 18 with PD and 24.2 million people over the age of 65 [10]. In Mexico, the prevalence is 19.9% [11]. PD increases the risk of type 2 diabetes, stroke, and myocardial infarction (MI) by 3- to 10-fold [12].

1.1. Adipose Tissue

Adipose tissue (AT) is a complex organ composed of adipocytes, stromal cells (macrophages), and endothelial and blood cells [13]. Based on its diverse characteristics, it is classified into four types: white adipose tissue (WAT), pink adipose tissue (PAT), beige adipose tissue (BAT), and brown adipose tissue (BBT) [14]. Depending on its location, adipose tissue is classified as visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) [15,16].

Visceral adipose tissue (VAT) is the fat adjacent to intra-abdominal organs and is the cornerstone of the etiology of type 2 diabetes, prediabetes, and cardiovascular disease (CVD) [14]. When diet exceeds energy expenditure, VAT accumulates in the visceral depot and induces free fatty acid (FFA) lipotoxicity and decreased insulin sensitivity (IS) [16,17].

In normal individuals, insulin inhibits lipolysis, stimulates the uptake of glucose and circulating fatty acids into adipose tissue (AT), and promotes triglyceride (TG) synthesis by suppressing hormone-sensitive lipase [18]. Adipokines and free fatty acids secreted by AT cause apoptosis of pancreatic β cells and decrease insulin synthesis and secretion [19].

The Mexican population is metabolically distinct from other ethnic groups, as it has a high predisposition to early-onset overweight and obesity. In overweight and obese individuals, excess VAT causes alterations in various metabolic pathways. Initially, dephosphorylation and inactivation of the mitogen-activated protein kinase (MAPK) pathway occurs, inducing inflammation in adipose tissue and generating insulin resistance (IR) by directly inactivating insulin receptor substrate 1 (IRS1) and indirectly inactivating peroxisome proliferator-activated receptor ɣ (PPAR-ɣ). This leads to increased lipotoxicity and decreased peripheral insulin sensitivity.

1.2. Body Composition

Body fat can be measured by various methods (magnetic resonance imaging, dual-energy X-ray absorptiometry (DEXA), bioimpedance (BIA), etc.). It is measured by fractionating total body fat into different units.

In our study, we decided to use bioimpedance (BIA) because it is an indirect, noninvasive, and inexpensive method that estimates body composition and consists of measuring tissue resistance to the passage of alternating current. BIA has a high diagnostic accuracy (r² = 0.84–0.90 and a coefficient of variation of 1.7–4.5%). However, its accuracy is influenced by electrode placement, hydration level, diet, menstruation, and ambient temperature; therefore, proper preparation before the study is important [20,21,22,23].

1.3. Treatments for Obesity

Lifestyle changes are the cornerstone of obesity treatment and the first step in treating prediabetes and type 2 diabetes [14]. The American Association of Clinical Endocrinologists (AACE), the Endocrine Society Obesity Guidelines, and the latest consensus report from the European Association for the Study of Diabetes/ADA recommend that the effect on weight be considered when choosing treatment for type 2 diabetes [24].

Glucagon-like peptide-1 (GLP-1) is a polypeptide incretin hormone that stimulates insulin secretion based on the amount of oral glucose consumed, lowers plasma glucose, reduces glucagon, slows gastric emptying, and suppresses appetite [14,25]. GLP-1 enhances insulin synthesis and secretion by binding to GLP-1 receptors on β cells, thereby increasing their sensitivity to glucose [26,27].

Liraglutide is a GLP-1 receptor analog (GLP-1RA) approved by the FDA in doses up to 1.8 mg in combination with insulin for the treatment of type 2 diabetes, and in a dose of 3.0 mg for overweight or obese individuals [24].

There are multiple meta-analyses demonstrating the efficacy of GLP-1 analogs in reducing VAT, weight loss, controlling glycated hemoglobin, controlling fasting glucose, and lowering high-density lipoprotein cholesterol (LDL-C) in patients with type 2 diabetes, in patients with prediabetes, and in patients with obesity; however, few studies compare VAT reduction in both groups (obesity and type 2 diabetes vs. obesity and prediabetes). Therefore, the following research question was posed: What is the difference in visceral adipose tissue quantification in people with obesity and type 2 diabetes vs. obesity and prediabetes treated with liraglutide in Mexico during 2023 and 2024?

The importance of our study lies in comparing which group (G1: OB + PD/G2: OB + T2D) presents a greater decrease in visceral adipose tissue, and therefore better metabolic control (improved glucose levels, lipid profile, glycosylated hemoglobin) due to the alterations discussed above.

The primary outcome was to compare the reduction in visceral adipose tissue in patients with obesity and prediabetes vs. obesity and type 2 diabetes treated with liraglutide. The secondary outcome was to describe metabolic control (glucose, A1C%, lipid profile), anthropometric changes (waist circumference and weight), and bioimpedance (total body fat, muscle mass).

2. Materials and Methods

With prior signature, authorization, and informed consent, a quasi-experimental, analytical, prolective, longitudinal study was conducted for a period of 6 months at a third-level center (Centro Medico Nacional La Raza “Dr. Antonio Fraga Mouret”) from 7 January 2023 to 3 January 2024. This study was approved by the institution’s ethics committee with folio R-2023-3501-093.

A study was conducted whose baseline status included men and women with OB, divided into two groups: G1 (OB and PD) vs. G2 (OB and T2D) according to the ADA definition. Using the mean difference formula with a 95% confidence level, a 1:1 ratio, and a statistical power of 80% with a standard deviation (SD) of 9.3 for group 1 and 12.3 for group 2 obtained from the article by Neeland et al. [14], a sample size of 16 participants per group was analyzed using the Open EPI program for difference in measurements. To this, 10% of the losses were added, obtaining a total of 36 participants (18 per group). Participants were recruited according to the inclusion, exclusion, non-exclusion, and elimination criteria described in the following section.

Follow-up was conducted at 0, 3, and 6 months. Anthropometric measurements (weight, height, BMI, waist circumference, total body fat, visceral fat, muscle tissue) and laboratory tests (glucose, A1c%, lipid profile) were performed at baseline, 3, and 6 months. Both groups received subcutaneous liraglutide. The medication was titrated weekly according to patient tolerance to reduce drug-associated side effects such as nausea, vomiting, diarrhea, or biliary pancreatitis. The target dose was 1.8 mg subcutaneously every 24 h, self-administered by the patient, which was confirmed by the patient returning the empty device to the research team. The target dose was reached by week 3 in 94.44% of patients, and 6.66% were escalated by week 5. To minimize the drug’s side effects, participants were given a personalized description of foods to avoid while taking the medication, permitted feeding techniques in case of significant oral intolerance, and information on emergencies suggestive of pancreatitis (transient epigastric abdominal pain radiating to the back, fever, nausea, vomiting, and tachycardia). In addition, an open appointment was made to the emergency room of this unit if any of these symptoms were present.

2.1. Inclusion Criteria

The inclusion criteria were men and women ≥ 18 years old, with type 2 diabetes, prediabetes, and obesity listed in NOM-008-SSA3-2010 [28].

2.2. Non-Inclusion Criteria

Non-inclusion criteria were pregnancy; type 1 diabetes; people with diseases that alter weight such as Cushing’s syndrome; poorly replaced hypothyroidism; chronic kidney disease or liver failure; elevated transaminases (≥5 times ULN) and/or known liver disease; biliary lithiasis; and use of hypoglycemic agents that alter weight (SGLT-2 inhibitors, sulfonylureas, insulins, and thiazolidinediones, without previous use of GLP-1 analog).

2.3. Exclusion Criteria

The exclusion criteria were men and women with loss of membership.

2.4. Elimination Criteria

The elimination criteria were men and women who do not apply the medication correctly and people who do not comply with the preparation for taking the bioimpedance.

Since our study center is a tertiary care hospital, most patients with type 2 diabetes use insulin and multiple oral hypoglycemic agents for glucose control. A search was conducted in primary care centers to find patients with type 2 diabetes who met the selection criteria. The Diabetes Care Centers of the Mexican Social Security Institute (CADIMSS) in Family Medicine Units (UMF) 19 and 15 were approached via in-person counseling and telephone. The CADIMSS centers specialize in training on injectable medication administration, carbohydrate counting, diet, exercise, and diabetic foot care in patients with recently diagnosed type 2 diabetes of the Mexican Social Security Institute (IMSS). Of a total of 210 participants, 152 were not evaluated; 104 patients were excluded (ineligible due to the use of injectable medications such as GLP-1 analogs or insulin and oral antidiabetics that affect weight, mentioned in the exclusion criteria section)—100 participants from CADIMSS and 4 participants from CMNR. Eligible but not recruited were 48 participants: 30 from CADIMSS were contacted by telephone but did not respond, and 18 participants from CMNR refused to participate in the protocol because they live in distant cities and are unable to travel every 3 months (our center, being a tertiary care center, receives participants from all over the country). A total of 58 participants were recruited. Of these 58, 10 were lost; 1 participant due to significant oral intolerance due to poor adherence to dietary recommendations, who was seen by the emergency department and referred to the gastroenterology service of our unit, currently stable; 5 participants were out of stock for more than 20 days (due to national health policies); 1 participant was diagnosed with eating disorders (which contraindicates the use of the molecule, and was already referred and treated by the psychology and psychiatry departments of this unit); and 3 participants lost social security benefits. The sample size was reduced to 26 participants. Therefore, a second recruitment of 10 more participants from CMNR was conducted. The total sample size was 36 participants, with 18 members in each group.

The statistical package SPSS v26.0 ^® and Microsoft Excel^® were used for data analysis.

2.5. Univariate Analysis

To determine the distribution of the variables, the Komolgorov–Smirnov test was performed. For quantitative variables with normal distribution, the results were reported as mean and standard deviation; for quantitative variables with free distribution, as median and interquartile ranges.

2.6. Bivariate Analysis

To demonstrate whether there is a difference in the decrease in VAT between both groups (OB and PD vs. OB and T2D) in the variables with normal distribution, the Student’s t-test was performed, and in variables with free distribution, the Mann–Whitney U test. For the intragroup or related samples analysis of VAT at 0, 3, and 6 months, a 1-factor ANOVA was performed for variables with normal distribution and Friedman for variables with free distribution.

2.7. Multivariate Analysis

No multivariate study was performed, since no statistical significance was found in VAT between both groups (OB and PD vs. OB and DT2).

2.8. Control for Bias

To avoid inappropriate set bias, the diagnostic criteria for T2D and PD established by the ADA were used, as was the operational definition of Obesity according to the official Mexican standard NOM-008-SSA3-2010 for the comprehensive treatment of overweight and obesity. To avoid susceptibility bias, patients with weight-modifying oral antidiabetics and patients who had previously used GLP-1 analogs were eliminated. To avoid performance bias and ensure medication administration, participants were asked to return the empty medication containers, and this was corroborated by weight loss and glycemic control through anthropometry and laboratory tests. To avoid detection bias and to prevent carryover bias from increasing the study population by 10%, the same number of measurements were taken in both groups.

2.9. Sample Maintenance

All patients were referred to psychology, psychiatry, and nutrition as part of our hospital protocol for subjects living with obesity to receive dietary guidance and rule out any eating disorder. For sample maintenance, participants were followed-up every 3 months.

2.10. Quality of the Maneuver

To avoid any bias in execution, body fat distribution was measured by bioimpedance according to established international preparation standards: fasting for 2 to 4 h prior to the measurement, removal of metal objects, no intense physical exercise 8 to 12 h prior, limbs in abduction and separated from the trunk, and the use of electrodes outside of injured areas and prior cleaning with 70° alcohol.

2.11. Transfer Control

To avoid transfer bias due to side effects of the medications, the sample size was increased to 20% estimated in the permissible range of losses.

2.12. Type of Analysis

An analysis was performed by protocol.

2.13. Ethical Considerations

This study was conducted in accordance with the Regulations of the General Health Law, articles 100, 101, 102, and 103; based on the Declaration of Helsinki, this research protocol’s degree of risk according to the General Health Law is considered research with greater than minimum risk. To minimize the side effects of the medication, escalated doses of the medication were administered according to the patient’s tolerance, personalized description of the foods that should be avoided with the use of the medication, feeding techniques allowed in case of significant intolerance to the oral route, as well as emergency data suggestive of pancreatitis (transient abdominal pain in the epigastrium that radiates to the back, fever, nausea, vomiting, and tachycardia).

3. Results

Of the 36 participants, 36.1% (n = 13) were male and 63.9% (n = 23) were female. The mean age was 48.36 ± 10.85 years; 33.3% (n = 12) had grade 1 obesity, 15.0% (n = 15) grade 2 obesity, and 9% (n = 9) grade 3 obesity. Regarding anthropometric measurements, the average weight was 99.28 ± 19.39 kilos, height 1.65 ± 0.08 m, waist circumference 115.75 ± 12.72 cm, total body fat 43.90 ± 9.76%, muscle 24.72 ± 4.49%, and visceral fat 14.5 (IQR 8.0–30.0). Regarding the laboratory parameters, the median glucose was 6.25 mmol/L (IQR 3.55–16.44), glycated hemoglobin 6.05% (IQR 4.9–12.3), triglycerides 1.90 mmol/L (IQR 0.80–6.89), high-density cholesterol 1.03 mmol/L (IQR 0.71–1.66), average low-density cholesterol 2.58 ± 0.82 mmol/L, and total cholesterol 44.30 ± 11.01 mmol/L; statistically significant were the male sex, glucose, and glycated hemoglobin with a p = 0.053, 0.000, and 0.0002, respectively; as specified in

Table 1. Sociodemographic, clinical, and laboratory characteristics of the population.

Variable	Total n = 36
Sociodemographic variables
Age, years 1	48.36 ± 10.85
Sex
Men, n (%)	13 (36.1%)
Women, n (%)	23 (63.9%)
Anthropometry
Height, mts 1	1.65 ± 0.08
Weight, kg 1	99.28 ± 19.39
Waist circumference, cm 1	115.75 ± 12.72
BMI, kg/m2
• Grade 1 obesity (30–34.9)	12 (33.3%)
• Grade 2 obesity (35.0–39.9)	15.0 (41.7%)
• Grade 3 obesity (>40)	9.0 (25.0%)
Bioimpedance
Body fat, % 1	43.90 ± 9.76
Muscle, % 1	24.72 ± 4.49
Visceral fat, % 1	15.20 ± 5.91
Laboratory
Glucose, mmol/L 2	6.25 (3.55–16.44)
Glycated hemoglobin (A1c), % 2	6.05 (5.7–12.3)
Cholesterol, mmol/L 1	44.30 ± 11.01
Triglycerides, mmol/L 2	1.90 (0.80–6.89)
High-density lipoprotein (HDL) cholesterol, mmol/L 2	1.03 (0.71–1.66)
Low-density lipoprotein (LDL) cholesterol, mmol/L 1	2.58 ± 0.82

The information is presented in frequency (percentage), otherwise it is indicated as ¹ mean ± standard deviation or ² median or interquartile range, according to the distribution estimated with the Komolgorov–Smirnov test.

.

In G1 (OB and PD) 38.9% (n = 7) were male, 61.1% (n = 11) female, the average age was 48.36 ± 10.85 years, 38.9% (n = 7) showed grade 1 obesity, 16.67% (n = 3) grade 2 obesity, and 4.44% (n = 8) grade 3 obesity. In anthropometric measurements, mean height was 1.64 ± 0.08 m, weight 105.21 ± 20.94 kilos, waist circumference 119.22 ± 15.20 cm, body fat 44.83 ± 8.6%, muscle 24.58 ± 4.16%, and median visceral fat 16.0% (IQR 8.0–30.0). Of the biochemical parameters, the median glucose was 5.50 mmol/L (IQR 3.55–6.68), glycosylated hemoglobin 5.9% (IQR 5.7–6.2), triglycerides 1.76 mmol/L (IQR 0.89–6.89), high-density cholesterol 1.01 mmol/L (IQR 0.71–1.29), total cholesterol 44.01 ± 10.67 mmol/L, and low-density cholesterol 2.30 ± 350.90 mmol/L, as specified in

Table 2. Sociodemographic, clinical, and laboratory characteristics by intervention group.

Variable	Total n = 36	Obesity and Prediabetes n = 18	Obesity and Type 2 Diabetes n = 18	p
Sociodemographic variables
Age, years 1	48.36 ± 10.85	50.06 ± 9.32	46.67 ± 12.23	0.356 ᵻ
Sex 2				0.053 ≠
Men, n (%)	13.00 (36.1%)	7.00 (38.9%)	6.00 (33.3%)
Women, n (%)	23.00 (63.9%)	11.0 (61.1%)	12.0 (66.7%)
Time of evolution of diabetes 2				0.968 ≠
<5 years			12.00 (66.6%)
5 a 10 years			4.00 (22.2%)
>10 years			2.00 (11.1%)
Anthropometry
Height, m, mts 1	1.65 ± 0.08	1.64 ± 0.08	1.65 ± 0.08	0.582 ᵻ
Weight, kg 1	99.28 ± 19.39	105.21 ± 20.94	98.10 ± 11.80	0.280 ᵻ
Waist circumference, cm 1	115.75 ± 12.72	119.22 ± 15.20	112.28 ± 8.76	0.102 ᵻ
BMI, kg/m2				0.222 ≠
Grade 1 obesity (30–34.9)	12 (33.3%)	7.0 (38.9%)	5.0 (29.4%)
Grade 2 obesity (35.0–39.9)	15.0 (41.7%)	3.0 (16.67%)	12.0 (70.6%)
Grade 3 obesity (>40)	9.0 (25.0%)	8.0 (4.44%)	1 (5.56%)
Bioimpedance
Body fat, % 1	43.90 ± 9.76	44.83 ± 8.60	44.67 ± 8.91	0.956 ᵻ
Muscle, % 1	24.72 ± 4.49	24.58 ± 4.16	24.87 ±4.92	0.850 ᵻ
Visceral fat, % 1	15.20 ± 5.91	17.00 ±7.18	13.41 ± 3.69	0.009 ᵻ
Laboratory
Glucose, mmol/L 2	6.25 (3.55–16.44)	5.5 (3.55–6.68)	7.68 (4.44–16.44)	<0.005 *§
Glycated hemoglobin (A1c), % 2	6.05 (4.9–12.3)	5.9 (5.7–6.20)	7.60 (5.40–12.30)	<0.005 *§
Cholesterol, mmol/L 1	44.30 ± 11.01	44.01 ± 10.67	44.65 ± 10.68	0.870 ᵻ
Triglycerides, mmol/L 2	1.90 (0.80–6.89)	1.76 (0.98–6.89)	2.33 (0.80–6.05)	0.401 §
High-density lipoprotein (HDL) cholesterol, mmol/L 2	1.03 (0.71–1.66)	1.01 (0.71–1.29)	1.03 (0.78–1.66)	0.407 §
Low-density lipoprotein (LDL) cholesterol, mg/dL 1	2.58 ± 0.82	2.30 ± 0.90	2.86 ± 0.65	<0.042 *ᵻ

The information is presented in frequency (percentage), otherwise it is indicated as ¹ mean ± standard deviation or ² median (interquartile range), according to the distribution estimated with the Komolgorov–Smirnov test. The ^≠ Chi-square test was used for dichotomous variables; ^ᵻ: Student’s t-test for the difference in means in quantitative variables with a normal distribution; ^§: Mann–Whitney U test for the difference in means in quantitative variables with a free distribution, Kruskal–Wallis for ordinal qualitative variables. * Significance is shown in bold and was set at a p value < 0.05.

.

In G2 (OB and T2D, 33.3% (n = 6) were male and 66.7% (n = 12) were female. The average age was 46.67 ± 12.23 years, 66.6% (n = 12) had <5 years of diabetes diagnosis, 22.2% (n = 4) 5 to 10 years, and 11.1% (n = 2) > 10 years; regarding the degree of OB, 29.4% (n = 5) were in OB grade 1, 70.6% (n = 12) OB grade 2, and 5.56% (n = 1) obesity grade 3. Regarding anthropometric measurements, the average height was 1.65 ± 0.08 m, weight 98.10 ± 11.80 kilos, total body fat 44.67 ± 8.91%, muscle 24.87 ± 4.92%, and median visceral fat 12.5 (IQR 9.00–20.00). In the biochemical parameters, the median glucose was 7.68 mmol/L (IQR 4.44–16.44), glycosylated hemoglobin 7.6% (IQR 5.40–12.30), triglycerides 2.33 mmol/L (IQR 0.80–6.05), and high-density cholesterol 1.03 mmol/L (IQR 0.78–1.66); the total cholesterol was 44.65 ± 10.68 mmol/L, and low-density cholesterol 2.86 ± 0.65 mmol/L, as observed in Table 2.

3.1. Bivariate Analysis

3.1.1. Comparison Between Groups

When comparing groups G1 (OB and PD) vs. G2 (OB and T2D) at month 0, using Student’s t-test and Mann–Whitney U, statistical significance was found in glucose levels (−0.56 mmol/L; p = 0.002) and A1c% (−0.4%; p = 0.000), as well as at month 3 glucose (−1.36 mmol/L; p = 0.001) and A1c% (−0.75%; p = 0.000), and month 6 glucose (−0.99 mmol/L; p = 0.012) and A1c% (−1.05%; p = 0.000). At month 0, month 3, and month 6, no statistically significant differences were found in any anthropometric or bioimpedance parameters. Although significant differences were found in glucose and A1c% between both groups, these are not considered to have clinical value, since the baseline levels of each group were different according to the inclusion criteria.

3.1.2. Comparison Intragroup

When comparing intragroup (0-, 3-, and 6-month follow-up of G1 and G2), in G1 (OB and PD) statistical significance was found in weight (−3.4 kg; p = 0.000), waist circumference (−3.09 cm; p = 0.017), glucose (−0.56 mmol/L; p = 0.002), A1c% (−0.5%; p = 0.000), and HDL (+0.004 mmol/L; p = 0.049). In G2 (OB and T2D) statistical significance was found in weight (−3.78 kg; p = 0.012), waist circumference (−3.9%; p = 0.049), glucose (−1.75 mmol/L; p = 0.002), A1c% (−1.15%; p = 0.001), HDL-c (+0.27 mmol/L; p = 0.001), and LDL-c (−0.17 mmol/L; p = 0.004). At month 0, month 3, and month 6, no statistically significant differences were found in any bioimpedance parameters, as described in

Table 3. Anthropometric and biochemical changes in the obesity and prediabetes group vs. obesity and type 2 diabetes treated with liraglutide 1.8 mg subcutaneously every 24 h.

Analysis of Unrelated Samples (Between Groups)
Variable	Month 0 n = 36			Month 3 n = 36			Month 6 n = 36
	Obesity and Prediabetes n = 18	Obesity and Type 2 Diabetes n = 18	p	Obesity and Prediabetes n = 18	Obesity and Type 2 Diabetes n = 18	p	Obesity and Prediabetes n = 18	Obesity and Type 2 Diabetes n = 18	p
Anthropometric
Weight, kg 1	105.21 ± 20.94	98.10 ± 11.80	0.226	101.33 ± 22.31	95.77 ± 12.28	0.352 ᵻ	101.81 ± 23.89	94.32 ± 12.30	0.269 ᵻ
Waist circumference, cm 1	119.22 ± 15.20	112.28 ± 8.76	0.098	115.41 ± 16.33	110.56 ± 9.57	0.273 ᵻ	116.13 ± 17.64	108.38 ±8.90	0.129 ᵻ
BMI, kg/m2 2	37.30 (30.46–54.83)	35.77 (30.10–42.17)	0.401	35.56 (28.08–53.61)	34.69 (29.51–41.93)	0.673 §	37.65 (27.47–54.53)	34.05 (29.51–41.93)	0.573 §
Bioimpedance
Body fat, % 1	44.83 ± 8.60	44.67 ± 8.91	0.910	44.77 ± 8.81	42.38 ± 10.43	0.414 ᵻ	45.57 ± 7.82	43.53 ± 6.89	0.427 ᵻ
Muscle, % 1	24.58 ± 4.16	24.87 ± 4.92	0.820	24.28 ± 3.83	24.63 ± 3.91	0.760 ᵻ	24.48 ± 3.74	25.46 ± 3.97	0.463 ᵻ
Visceral fat, % 1	17.00 ± 7.18	13.41 ± 3.69	0.081	15.00 ± 6.69	12.42 ± 3.77	0.218 §	15.33 ± 7.20	14.00 ± 4.47	0.626 ᵻ
Laboratory
Glucose, mg/dL 2	5.50 (3.55–6.68)	7.68(4.44–16.44)	0.002 *	4.76 (4.04–6.15)	6.12 (3.83–9.78)	0.001 *	4.94 (3.99–6.21)	5.93 (4.13–8.24)	0.012 *§
Glycated hemoglobin (A1c), % 2	5.90 (5.70–6.30)	7.60 (5.7–12.30)	0.010 *	5.50 (4.9–6.0)	6.25 (42.20 ± 13.00	<0.005 *	5.40 (4.90–5.80)	6.45 (5.0–9.10)	0.000 *§
Cholesterol, mmol/L 1	44.01 ± 10.67	44.65 ± 10.68	0.949	43.63 ± 10.19	42.20 ± 13.00	0.715 ᵻ	43.57 ± 11.49	43.06 ± 10.90	0.541 §
Triglycerides, mmol/L 2	1.76 (0.98–6.89)	2.33 (0.80–6.05)	0.178	1.35 (0.72–10.65)	1.49 (0.83–5.00)	0.389 §	1.46 (0.49–7.58)	1.54 (0.74–6.04)	0.925 ᵻ
High-density lipoprotein (HDL) cholesterol, mmol/L 2	1.01 (0.71–1.29)	1.03 (0.78–1.66)	0.849	-	-	-	1.05 (0.81–1.60)	1.30 (0.71–1.76)	0.005 *§
Low-density lipoprotein (LDL) cholesterol, mmol/L 1	2.30 ± 0.90	2.86 ± 0.65	0.105	-	-	0.659 ᵻ	2.01 ± 0.53	2.69 ± 0.74	0.056 ᵻ
Analysis of related samples (intragroup)
Variable	Obesity and Prediabetes n = 18				Obesity and Type 2 Diabetes n = 18
	Month 0	Month 3	Month 6	p	Month 0	Month 3	Month 6	p
Anthropometric
Weight, kg 1	105.21 ± 20.94	101.33 ± 22.31	101.81 ± 23.89	0.000 *€	98.10 ± 11.80	95.77 ± 12.28	94.32 ± 12.30	0.012 *€
Waist circumference, cm 1	119.22 ± 15.20	115.41 ± 16.33	116.13 ± 17.64	0.017 *€	112.28 ± 8.76	110.56 ± 9.57	108.38 ± 8.90	0.049 *€
BMI, kg/m2 2	37.30 (30.46–54.83)	35.56 (28.08–53.61)	37.65 (27.47–54.53)	0.002 *©	37.30 (30.46–54.83)	35.56 (28.08–53.61)	37.65 (27.47–54.53)	0.002 *©
Bioimpedance
Body fat, % 1	44.83 ± 8.60	44.77 ± 8.81	44.77 ± 8.81	0.77 6 €	44.67 ± 8.91	42.38 ± 10.43	43.53 ± 6.89	0.389 €
Muscle, % 1	24.58 ± 4.16	24.28 ± 3.83	24.28 ± 3.83	0.771 €	24.87 ± 16.0	24.63 ± 3.91	25.46 ± 3.97	0.154 €
Visceral fat, % 1	17.00 ± 7.18	15.00 ± 6.69	15.00 ± 6.69	0.077 €	13.41 ± 3.69	12.42 ± 3.77	14.00 ± 4.47	0.107 €
Laboratory
Glucose, mmol/L 2	5.50 (3.55–6.68)	4.76 (4.04–6.15)	4.94 (3.99–6.21)	0.002 *©	7.68 (4.44–16.44)	6.12 (3.83–9.78)	5.93 (4.13–8.24)	0.002 *©
Glycated hemoglobin (A1c), % 2	5.90 (5.70–6.30)	5.50 (4.9–6.0)	5.40 (4.90–5.80)	0.000 *©	7.60 (5.40–12.13)	6.25 (5.5–9.5)	6.45 (5.0–9.10)	0.001 *©
Cholesterol, mmol/L 1	44.01 ± 10.67	43.63 ± 10.19	43.57 ± 11.49	0.969 €	44.65 ± 10.68	42.20 ± 13.00	43.06 ± 10.90	0.556 €
Triglycerides, mg/dL 2	1.76 (0.98–6.89)	1.35 (0.72–10.65)	1.46 (0.49–7.58)	0.906 ©	2.33 (0.80–6.05)	1.49 (0.83–5.00)	1.46 (0.49–7.58)	0.141 ©
High-density lipoprotein (HDL) cholesterol, mmol/L 2	1.01 (0.71–1.29)	-	1.05 (0.81–1.60)	0.049 *ɷ	1.03 (0.78–1.66)	-	1.30 (0.71–1.76)	0.001 *ɷ
Low-density lipoprotein (LDL) cholesterol, mmol/L 1	2.30 ± 0.90	-	2.01 ± 0.53	0.069 ᵻ	2.86 ± 0.65	-	2.69 ± 0.74	0.004 *€ᵻ

The information is presented in frequency (percentage), otherwise it is indicated as ¹ mean ± standard deviation or ² median (interquartile range), according to the distribution estimated with the Komolgorov–Smirnov test. Unrelated and related samples analysis: Student’s ^ᵻ t-test was used for the difference in means in quantitative variables with a normal distribution and the ^§ Mann–Whitney U test for the difference in means in quantitative variables with a free distribution. Related samples analysis: ^€ ANOVA and Student’s ^ᵻ t-test were used for the difference in means in quantitative variables with a normal distribution of 3 and 2 groups, respectively, and ^© Friedman and ^ɷ Wilxocon for the difference in means in quantitative variables with a free distribution of 3 and 2 groups, respectively. * Significance is shown in bold and was set at a p value <0.05.

.

Although no statistical significance was found in the change in visceral adipose tissue between G1 and G2, graphing the anthropometric and laboratory changes shows that G1 (OB + PD) had a more pronounced weight loss in the first 3 months, with a subsequent plateau; this was in contrast to G2 (OB + T2D), who had a more consistent weight loss over the 6 months. G2 (OB + T2D) had a greater loss of total body fat, followed by a faster recovery of muscle mass compared to G1 (OB + PD), as seen in Figure 1.

Regarding the biochemical results, G2 (OB + T2D) had greater glucose control, greater reductions in glucose and A1C%, and elevation of LCL-C were observed in Figure 1.

In addition, the main side effects presented by the participants were recorded. The most frequent were gastrointestinal in 27.8% of participants, nausea in 19.4%, diarrhea in 27.8%, vomiting in 8.33%, followed by headache in 11.11% and urticaria at the application site in 5.55%. One participant was excluded from the study due to a documented eating disorder with one sentinel event (vomiting coffee grounds secondary to a Mallory Weiss tear) who was treated in this unit by the gastroenterology and internal medicine service without complications. Upon discharge he was sent to the psychiatry service where he continued to be monitored. An X² test was performed to demonstrate the difference in means, without statistical significance or relative risk of importance.

4. Discussion

This study aimed to compare the reduction in visceral adipose tissue in patients with OB and PD versus OB and T2D, as there is no article in the literature comparing which group presents greater benefit.

4.1. Normal Function of Incretins

Native GLP-1 is a 30-amino acid peptide produced in intestinal epithelial L cells of the distal ileum and colon, and α cells of the pancreas and the CNS. It has a half-life of 1 to 2 min and is degraded by the proteolytic enzyme dipeptidyl dipeptidase-4 (DPP-4). GLP-1 is a hormone belonging to the incretin group, secreted by enteroendocrine L and α cells of the pancreas and the CNS. Its main function is to decrease β-cell apoptosis and increase neogenesis and proliferation [29]. Its function begins by binding to its G protein-coupled receptor (GLP-1R) and activates various metabolic pathways. Through the adenylate cyclase (AC) pathway, it increases cyclic adenosine monophosphate (cAMP); through protein kinase A (PKA), it promotes the exocytosis of insulin vesicles from the β cells of the pancreas; and finally, through the PI3K/mTOR pathway, it induces the activation of hypoxia-inducible factor 1α (HIF-1α), which activates glycolysis, the Krebs cycle, and increases intracellular adenosine triphosphate (ATP) [30]. This increase in ATP activates the closure of ATP-dependent potassium channels, increasing calcium channels whose gradient passage depolarizes the cell membrane and releases insulin [31]. Normally, this pathway is activated 10 min after the initial glucose peak, followed by a second phase of sustained release lasting up to 60 min [30,32]. In addition, it inhibits glucagon release from pancreatic α-cells, stimulates epidermal growth factor (EGF), the phosphatidylinositol-3 kinase (PI3K) pathway, and β-cell growth factor [29,31,32]. It also inhibits appetite by stimulating propiomelanocortin (POMC) and cocaine- and amphetamine-regulated transcriptase (CART), responsible for the anoxygenic pathway, and suppresses agouti-related neurons (AgRP) and neuropeptide Y (NPY), responsible for the oxygenic pathway, through ɣ-aminobutyric acid (GABA)-dependent signaling [33].

4.2. Alterations in the Incretin Effect in Prediabetes and Type 2 Diabetes

GLP-1 and Glucose-Dependent Insulin-Optic Polypeptide (GIP) are the major incretins secreted by the body (accounting for 90% of incretin activity) and generate 70 to 80% of insulin release after oral glucose administration. By binding to the GLP-1R, it activates the Epidermal Growth Factor (EGF), which activates phosphatidylinositol-3 kinase (PI3K-AKT) and increases the growth of pancreatic B cells. In addition, GLP-1 promotes the expression of the pancreatic and duodenal homeobox-1 gene (PDX-1), which increases the nuclear translocation of insulin, the Facilitative Glucose Transporters Type 2 (GLUT-2), and glucokinase, replenishing B cell reserves and preventing cellular exhaustion. In people with T2D, a 50% decrease in GLP-1 levels is observed compared to healthy individuals, and an elevation of GIP caused by IR is observed. In people with BP, there is impaired secretion/activity of incretins, mainly GLP-1, in the early phase. All of these metabolic alterations (T2D and PD) are often associated with weight due to excess VAT.

4.3. Metabolic Alterations in Visceral Adipose Tissue (VAT)

Excess nutrients, physical inactivity, and carbohydrate-rich meals are often risk factors that increase VAT. It is known that the accumulation of VAT generates greater lipolysis and lower SI due to a greater amount of FFA in the form of TGC [17]. Excess VAT produces alterations in different metabolic pathways.

• Dephosphorylation and inactivation of the MAP kinase (MAPK) pathway: This induces adipose tissue inflammation and insulin resistance (IR) by indirectly inactivating insulin receptor substrate 1 (IRS1) and peroxisome proliferator-activated receptor ɣ (PPAR-ɣ).
• Aberrant phosphatidylinositol-3 kinase (PI3K-AKT) pathway: PI3K transforms phosphatidylinositol 4,5-bisphosphate (PIP2) into phosphatidylinositol 3,4,5-trisphosphate (PIP3), which activates phosphoinositide-dependent kinases and AKT that regulate glycogen synthesis, glucose uptake, and adipogenesis. Excess VAT produces a selective inhibition of the PI3K pathway, which eliminates the effect of leptin (suppressing food intake in the hypothalamus) [33,34]. Furthermore, inhibition of the PI3K/AKT pathway degrades Sort1, a key element in the storage of glucose transporter 4 (GLUT4), decreasing insulin sensitivity [35].
• Janus kinase (JAK)/Signal transducer and activator of transcription (STAT) pathway suppression: In normal-weight individuals, the pathway activates propiomelanocortin (POMC) which suppresses food intake. In excess VAT, there is a suppression of the JAK/STAT pathway in the central nervous system (CNS) which reduces leptin sensitivity in POMC neurons. Furthermore, hepatic steatosis is caused by an aberration in the JAK/STAT pathway mediated by growth factors and cytokines, which is associated with IR and increased expression of gluconeogenesis genes [36].
• Elevated transforming growth factor β (TGF-β): It has a dual effect on adipogenesis and adipocyte differentiation. It inhibits mesenchymal stem cell (MSC) differentiation by phosphorylating and suppressing PPAR-ɣ, which increases adipose cell expansion in the bone marrow [37]. Excessive VAT presents elevated levels of TGF-β, and aerobic exercise suppresses these levels [38].

4.4. GLP-1 Receptor Analogs

The Mexican Obesity Association (SMO) recommends that patients with glucose disorders (carbohydrate intolerance, dyslipidemia) prefer medications that reduce weight and VAT for greater metabolic benefit. The ADA recommends considering the effect of weight when choosing a hypoglycemic therapy for diabetes. There are medications approved for weight loss and glucose control in patients with PD and T2D that belong to the group of medications called GLP-1 analogs.

Liraglutide is a GLP-1 receptor, selected in our study due to its high homology with human GLP-1 (97%). It has an arginine–lysine substitution at position 34 and a fatty acid residue at lysine 26; this modification allows for a drug half-life of 11 to 13 h [29,30]. This modification allows for the prolongation of the effects of GLP-1 receptors while maintaining insulin production through the mTOR-dependent HIF-1α pathway [39]. The FDA approved its use at doses up to 1.8 mg in patients with T2DM and 3.0 mg in patients with obesity.

In our study, we used a dose of 1.8 mg, the maximum allowed by our institution for both groups. GLP-1RAs are involved in satiety, thermogenesis, blood pressure, the reduction of chronic inflammation, and neurogenesis as part of CNS processes.

GLP-1R is found in macrophages, lymphocytes, and monocytes and suppresses tumor necrosis factor α (TNF-α), interleukin 6 (IL-6), and interleukin 1β (IL-1β), thereby decreasing oxidative stress, chronic inflammation, and improving endothelial function [40]. At the brain level, they are found in the cerebral cortex, thalamus, hypothalamus, substantia nigra, and cerebellum, crossing the blood–brain barrier and exerting control over satiety and hunger [30]. At the gastrointestinal level, they decrease food consumption, slow gastric emptying, and promote the release of hormones such as leptin and peptide YY [41]. In adipose tissue, they stimulate the WNT/B-catenin pathway that inhibits de novo lipogenesis (genes such as DGAT1, SCD1, Apo B, FABP1, and FOXA1) involved in the synthesis of fatty acids and TGC [42]. In lipid metabolism, they stimulate the cAMP pathway, which activates Sirtulin 1 (SIRT1), a nicotianamine adenine dinucleotide (NAD)-dependent deacetylase that increases triacylglycerol lipase, which depletes TGCs in WAT, resulting in improved fat accumulation and energy expenditure, and inhibits PPAR-ɣ, which decreases proteins associated with lipid metabolism [42,43].

SCALE Diabetes, a 56-week, randomized, double-blind, multinational, multicenter, placebo-controlled trial of liraglutide at doses of 1.8 mg and 3.0 mg in people with type 2 diabetes and overweight or obesity treated with ≤2 oral antidiabetic agents (OADs), observed weight reductions of 4.7% and 6.0% with liraglutide at doses of 1.8 mg and 3.0 mg, respectively, compared with 2.0% with placebo [28]. In our study, group G1 (OB and type 2 diabetes mellitus) had less weight loss (−3.4 kg; p = 0.000) compared to group G2 (OB and type 2 diabetes mellitus) (−3.78 kg; p = 0.012) [44].

Santilli, F. et al. conducted a longitudinal, randomized, controlled, parallel-arm study in OB patients treated with metformin and PD or newly diagnosed type 2 diabetes (<1 year since diagnosis) receiving liraglutide (1.8 mg/day) versus lifestyle changes. They found a greater reduction in VAT in the liraglutide group (−15.3% vs. −9.0%) with no statistical differences in SBP [25]. Neeland, I. et al, conducted a randomized, placebo-controlled clinical trial with liraglutide at a dose of 3.0 mg to evaluate the change in visceral and ectopic fat in people with PB and SP without diabetes and high CVR. They found that liraglutide achieved a two-fold greater reduction in abdominal visceral fat and a six-fold greater reduction in liver fat (−12.49% with liraglutide vs. −1.63% with placebo) [14].

Our study focused on patients with weight disorders and glucose problems secondary to elevated VAT levels. Among the results found, 100% of patients presented clinical manifestations of IR (acanthosis nigricans and presence of skin tags), which was corroborated by measuring a 75 g oral glucose tolerance test (OGTT), fasting glucose, and A1c%. Both groups received liraglutide at a dose of 1.8 mg/day. The intragroup comparison showed statistical significance in weight G2 (OB and T2D) (−3.78 kg; p = 0.012) vs. G1 (OB and T2D) (−3.78 kg; p = 0.012), waist circumference G1 (OB and T2D) (−3.9 cm; p = 0.049) vs. G1 (OB and PD) (−3.09 mmol/L; p = 0.017), glucose G2 (OB and T2D) (−1.75 mmol/L; p = 0.002) vs. G1 (OB and PD) (−0.56 mmol/L; p = 0.002), A1c% G2 (OB and T2D) (−1.15%; p = 0.001) vs. G1 (OB and PD) (−0.5%; p = 0.000), and HDL-c G2 (OB and T2D) (+0.27 mmol/L; p = 0.001) vs. G1 (OB and PD) (+0.004 mmol/L; p < 0.049), respectively. Greater weight loss, glycemic control, reduction in A1c percentage, and greater increase in HDL-c were observed in G2 (OB and T2D). At months 0, 3, and 6, no differences were observed in the mean value of the mean difference in A1c percentage and a greater increase in HDL-c in patients with type 2 diabetes. No statistically significant differences were found in any bioimpedance parameter, as described in Table 3. However, it is worth noting that G1 (OB and PD) presented more accelerated weight loss, metabolic control, and VAT reduction in the first 3 months, with a decrease from 3 to 6 months later. This was in contrast to G2 (OB and T2D), whose weight loss, metabolic control, and VAT reduction were slower, more constant, and sustained (at 6 months of follow-up), as evidenced in Figure 2. It is likely that the group that benefited the most is G2 (OB and T2D), because T2D has a 50% loss of GLP-1 function, and the application of this medication (GLP-1RA) improves the expression of GLUT-4 receptors, the MAP kinase pathway, decreases TGF-β, reactivates the PPAR-ɣ pathway, and decreases chronic inflammation by inhibiting TNF-α, IL-6, and IL-1β, as well as reactivating appetite.

5. Conclusions

Treatment with liraglutide offers benefits to both the population with obesity and prediabetes and obesity and type 2 diabetes, since this drug is a GLP-1RA. In obesity, prediabetes, and type 2 diabetes, excess VAT generates important metabolic alterations such as insulin resistance, glucolipotoxicity, and increased cardiovascular risk. The pathophysiology of excess VAT leads to dysfunction in the pancreatic β-cell. GLP-1R analogs are molecules that partially restore the incretin effect by decreasing VAT. Our study aimed to compare the decrease in VAT in patients with OB + PD and OB + T2D to determine which population benefited the most. No statistically significant differences were found in anthropometric measurements (weight, height, BMI) between G1 (OB + PD) and G2 (OB + T2D), nor in bioimpedance (visceral adipose tissue, subcutaneous adipose tissue, and muscle). However, when performing intragroup comparisons (each group at 0, 3, and 6 months), greater weight loss, decreased waist circumference, and better metabolic control (glucose, A1c%, HDL-C, and LDL-C) were found in G2 (OB and T2D), which remained constant throughout the 6-month follow-up. This indicates a greater benefit for this population group, since G1 (OB and PD) presented greater weight loss and total body fat loss in the first 3 months, with a plateau from month 3 to 6 of follow-up.